Table 2 Statistical approaches summarized according to the statistical challenge handled
From: The statistical approach in trial-based economic evaluations matters: get your statistics together!
Analysis | Baseline imbalances | Skewed costs | Correlated costs and effects | Missing data |
|---|---|---|---|---|
1.- Two t-tests (1 for costs; 1 for effects) complete-case analysis | ✗ | ✗ | ✗ | ✗ |
2.- Two bootstrapped t-tests; complete-case analysis | ✗ | ✓ | ✗ | ✗ |
3.- Two regressions with correction for confounders; complete-case analysis | ✓ | ✗ | ✗ | ✗ |
4.- Two bootstrapped regressions with correction for confounders; complete-case analysis | ✓ | ✓ | ✗ | ✗ |
5.- Two t-tests; mean imputation | ✗ | ✗ | ✗ | ✓ |
6.- Two bootstrapped t-tests; mean imputation | ✗ | ✓ | ✗ | ✓ |
7.- Two regressions with correction for confounders; mean imputation | ✓ | ✗ | ✗ | ✓ |
8.- Two bootstrapped regressions with correction for confounders; mean imputation | ✓ | ✓ | ✗ | ✓ |
9.- Two t-tests; MI | ✗ | ✗ | ✗ | ✓ |
10.- Two bootstrapped t-tests; MI | ✗ | ✓ | ✗ | ✓ |
11.- Two regressions without correction for confounders; MI | ✓ | ✗ | ✗ | ✓ |
12.- Two bootstrapped regression with correction for confounders; MI | ✓ | ✓ | ✗ | ✓ |
13.- SUR without correction for confounders; MI | ✗ | ✓ | ✓ | ✓ |
14.- SUR with correction for confounders; MI | ✓ | ✓ | ✓ | ✓ |